NEW YORK – In the several years since its launch, Bruker's timsTOF Pro mass spectrometer has made significant inroads into the proteomics market with a number of leading labs and companies developing workflows and software for use on the platform.
Following this trend, researchers from the University of Michigan recently detailed a new software package they said will greatly speed data analysis times while improving protein identification and quantification on the system.
Described in a study published this month in Molecular & Cellular Proteomics, the tool is an updated version of the MSFragger software developed in the lab of Alexey Nesvizhskii, professor of computation medicine and bioinformatics at Michigan and senior author on the MCP paper.
According to Nesvizhskii, the software is aimed in particular at improving analysis of data generated using the instrument's PASEF (Parallel Accumulation - Serial Fragmentation) workflow, which a number of existing software tools have struggled to deal with.
The PASEF method, which was developed by the lab of Max Planck Institute of Biochemistry researcher Matthias Mann, combines collection of ions via the timsTOF Pro's trapped ion mobility system (TIMS) with rapid quadrupole switching on a QTOF instrument to enable the fragmentation of multiple simultaneously eluting precursor ions.
In PASEF, precursor ions are accumulated in the TIMS device and then released in discrete packets based on their collisional cross sections. Then, instead of looking at only one set of precursors per scan, the instrument's quadrupole is set to rapidly switch from one precursor to another, which lets the mass spec analyze multiple precursors per scan and increases the speed of the MS/MS analysis by the number of precursors multiplexed.
This approach allows the instrument to collect large amounts of data quickly, but, Nesvizhskii noted, this data then needs to be analyzed, which has proved a bottleneck for some.
"You have the fourth dimension [of data] from the ion mobility separation, so the data are inherently more complex," he said. The timsTOF Pro "is also extremely fast. So on top of that increased complexity you also have a lot of mass spec scans, so a lot more data."
This means that "even if you want to do a simple analysis, it will take longer to search the data," Nesvizhskii said, noting that this has created situations where a two-hour mass spec run then requires upwards of 10 hours to analyze the data.
"That's really not good," he said. "You want to be able to analyze the data in less time than it took you to acquire it."
Brett Phinney, head of the proteomics core at the University of California, Davis, said he has firsthand experience with the sort of issue Nesvizhskii described.
"When we first got [the timsTOF Pro] I was kind of pulling my hair out because everything would take so long to process," he said.
To address the problem, Nesvizhskii and his colleagues adapted their MSFragger software to support timsTOF PASEF data. Originally published in 2017, MSFragger uses an indexing scheme that organizes both precursor and fragment peptides in such a way that it scores an experimental spectrum against all theoretical peptides simultaneously, which allows the software to match experimental and reference spectra more quickly than conventional approaches.
The researchers originally developed the tool to enable open database searching, in which matches to experimental spectrum are searched not against a discrete reference database but rather against the full range of theoretically possible peptides, allowing the detection of peptide forms like various post-translationally modified molecules not present in typical reference databases.
The software's emphasis on speed also makes it well-suited to more speedily analyzing timsTOF PASEF data. The researchers also adapted the software to read raw data from the Bruker platform directly, which eliminates the need to convert the data to a searchable format, which Nesvizhskii noted can also take an hour or more.
"We were excited about the [timsTOF Pro] because it is always good when there is competition in the field between different mass spec companies, so we wanted to make sure that if there was a new great instrument on the market that people could analyze the data," he said. "And we felt that with MSFragger and the unique advantages that it had, that with an instrument like timsTOF that is very fast and generates a lot of challenging data, we would be in a position to provide a computation tool that would help people."
The package also features a new tool for label-free quantification called IonQuant, which is meant to speed extraction of quantitative information from the LC-IMS-MS data.
In an analysis of a HeLa dataset, the researchers found that they could identify roughly 30 percent more unique peptides than the widely used MaxQuant software and around 10 percent more than the PEAKS software, with IonQuant quantifying more proteins than either. The MCP study found that the software was able to process a two-hour PASEF run in under 70 minutes on a desktop computer.
Phinney said that after his initial struggles he switched to the MSFragger software for his timsTOF Pro data and has had no problem with analysis times since. He noted that data processing and analysis times for the instrument were now on par with the times for the Thermo Fisher Scientific instruments he used.
"Everything is pretty much solved as far as I'm concerned," he said. "The software is as fast as I need it to be now, so it's not a problem anymore, but at the beginning it was really painful."
He said that in his experience this was not uncommon for new instruments following their release.
Phinney added that in addition to MSFragger, several other software packages he used appeared to work well with the timsTOF data, including the University of Washington's Skyline software and Biognosys's Spectronaut software.